Does periodization matter? The effect of different high intensity periodization models on endurance adaptations.

Transcription

1 INTRODUCTION To maximize physiological adaptations and performance capability in elite athletes, all factors involved in the training organization need to be optimized. In endurance sports, these include the duration and intensity of individual training sessions, the frequency of training sessions, and the organizational pattern of these stimulus variables over time. Recent descriptive studies of some of the world s best endurance athletes have shown that successful athletes in cycling (13, 26, 37), running (1, 2) and cross-country (XC) skiing (22, 23, 35) perform a high volume of low intensity training (LIT) (defined as work eliciting a stable blood lactate concentration [la - ] of less than approximately 2 mmol. L -1 ) in addition to much smaller but substantial proportions of both moderate intensity training (MIT) (2-4 mmol. L -1 blood lactate) and high intensity training (HIT) (training above maximum lactate steady-state intensity (>4 mmol. L -1 blood lactate)) throughout the preparation period. The majority of descriptive studies present a pyramidal training intensity distribution (TID), with high volume of LIT, substantial MIT and less HIT, while a few studies suggest athletes to adopt a polarized TID (reduced volume of MIT, somewhat higher HIT) which have been proposed to give superior endurance adaptions (29, 31). However, although some evidence suggests superior responses by increased HIT in a clearly polarized TID, there is currently limited empirical data comparing different stimulus ordering approaches for the HIT component of training that is often seen as critical to maximizing adaptations. The term training periodization originates primarily from older eastern European texts and is widely and rather indiscriminately used to describe and quantify the planning process of training (16). Periodization plans add training load-structure, with well-defined training periods designed to stimulate specific physiological adaptations (e.g. O 2max ) or performance qualities in a specific order presumed optimal for performance development. Such endurance training models involve manipulation of different training sessions periodized over timescales ranging from micro- (2-7 days), to meso- (3-6 wk) and macro cycles (6-12 months; including preparation, competition and transition periods). Recent experimental findings indicate improved training adaptations following shorter, highly focused training periods of HIT compared to mixed programs with the same total quantity of intensive sessions (19-21). For example, Rønnestad (19) found superior effects of a 12-wk block periodization program, where each 4-wk cycle consisted of one wk of five HIT sessions, followed by three wk of one HIT session. wk -1, when compared to a traditional program incorporating two weekly HIT sessions. However, others report superior effects following a polarized TID compared to a HIT block periodized training concept (30). The latter study was, however, not conducted with groups performing the same quantity of HIT sessions, which may have affected the results. These recent findings confirm HIT to be an important stimulus for endurance adaptations, but also highlight mesocycle organization as a potential modifier of the adaptive response. Previous research has shown that the physiological adaptations to HIT sessions are also sensitive to the interactive effects of intensity and accumulated duration. For example, both Seiler et al. (28) and Sandbakk et al. (24) have recently demonstrated that slight reductions in HIT work intensity facilitated large increases in tolerable accumulated duration, and better overall adaptive responses in well-trained cyclists and cross-country skiers. While research has progressed our understanding of the intensity/accumulated duration relationship during HIT sessions and its relation to endurance performance development in an isolated fashion (24, 28), the accumulative effects of the order of such sessions are not well understood. Different patterns of HIT ordering are used by elite athletes. Some endurance athletes increase HIT intensity and decreasing HIT duration from the preparation to the competition period (34, 35). However, anecdotal evidence also shows that some successful athletes utilize a reversed model, where HIT intensity is decreased and HIT duration increased, or a mixed model with larger micro-variation of various HIT sessions (e.g. interval sessions) throughout the training period. Therefore, the main purpose of this study was to compare the effects of three different HIT models, balanced for total load but periodized in a specific mesocycle order or in a mixed distribution, on endurance adaptations during a 12-wk training period in well-trained endurance athletes. We simulated a preparation period in which athletes in Increasing (INC), Decreasing (DEC) and Mixed (MIX) HIT

2 groups performed training periods that were matched for all features (frequency, total volume, and overall HIT load) except the mesocycle order or distribution of HIT sessions. We hypothesized that the INC HIT organization would be best tolerated and give best overall adaptive effects. METHODS This was a multicenter study, involving three test centers completing the same controlled experimental trial. At each test center, three matched periodization groups were instructed to follow a 12-wk highvolume LIT model, in addition to a significant portion HIT performed as prescribed and supervised interval sessions. Performance and physiological tests were compared before and after the intervention period. Subjects Sixty-nine male cyclists (38±8 yr, O 2peak 62±6 ml. kg -1. min -1 ) were recruited to the study using announcements in social-media and through local cycling clubs. Inclusion criteria were: (1) male, (2) O 2peak >55 ml. kg -1. min -1, (3) training frequency >4 sessions. wk -1, (4) cycling experience >3 yr, (5) regularly competing, and (6) absence of known disease or exercise limitations. Study participation was administered from three different test-locations, including 29, 20 and 20 subjects, respectively. All subjects were categorized as well-trained (11) or at performance level 4 according to athlete categorization by DePauw et al. (6). All subjects completed the intervention. However, we excluded six subjects from the final analyses due to absence from post-testing, and/or <70% compliance with prescribed interval sessions. Excluded subjects were from MIX (2 subjects) and DEC (4 subjects) groups. The study was approved by the ethics committee of the Faculty for Health and Sport Science, University of Agder, and registered with the Norwegian Social Science Data Services (NSD). All subjects gave their verbal and written informed consent prior to study participation. Pre-intervention period Prior to intervention, a 6-wk pre-intervention period (PIP) was conducted to familiarize subjects with interval sessions included in the intervention period and with testing protocols (Figure 1). During the PIP, subjects were instructed to perform only one interval session each wk, combined with freely chosen (ad libitum) LIT volume. All subjects completed a questionnaire regarding training history the previous year, years of cycling experience, previous peak performance level and previous/current injuries and diseases. Pre-testing was performed at the end of the PIP (mid-december), and subjects were thereafter randomized into one of three different training groups (INC, DEC and MIX) matched for (1) age, (2) cycling experience and (3) O 2peak. Intervention period Training organization The training intervention was performed from early January to the end of March (12-wk), corresponding to the early preparation period for these cyclists and consisted of three, 4-wk mesocycles. Subjects were instructed to follow a mesocycle wk load structure as follows; wk 1; medium LIT volume and two supervised interval sessions, wk 2 and 3; high LIT volume and three supervised interval sessions, wk 4; reduced LIT volume by 50% compared to the previous two wk and one HIT session executed as a physiological test (results not presented). In total, each subject was prescribed 24 supervised interval sessions, in addition to laboratory testing, and self-organized ad libitum LIT equal to the subject s normal LIT volume. Each intervention group organized interval sessions in a specific periodized mesocycle order or in a mixed distribution during mesocycle 1-3 (Figure 1).

3 FIGURE 1: Study protocol. A 6-wk pre-intervention period, including familiarization to interval sessions, pre-testing and randomization (R), was followed by a 12-wk intervention period divided in three 4-wk mesocycles with different interval session prescriptions for each training group. All groups performed 24 supervised interval sessions, in addition to testing and ad libitum low intensity training. Increasing HIT (INC) group (n=23) performed 8 interval sessions as 4x16-min in mesocycle 1 (wk 1-4), 8 interval sessions as 4x8-min in mesocycle 2 (wk 5-8) and 8 interval sessions as 4x4-min in mesocycle 3 (wk 9-12). Decreasing HIT (DEC) group (n=20) performed interval sessions in the opposite mesocycle order as INC, and Mixed HIT (MIX) group (n=20) organized all 24 interval sessions (8 in each mesocycle) in a mixed distribution; sessions 1 as 4x16 min, session 2 as 4x8 min, session 3 as 4x4 min, session 4 as 4x16 min and so on. In total during 12 wk, all subjects independent of group performed 8 interval sessions in each 4x16, 4x8 and 4x4 min prescriptions, respectively. All subjects were tested (T) in-between cycles during wk 4 and 8 (results not presented). Post-testing was completed within 5 days post intervention period. Interval sessions All HIT was performed indoors as supervised group interval training sessions, and included a min low-intensity (55-70% HR max ) warm up, followed by four interval bouts of 4, 8 or 16 min separated by 2 min rest, and concluded with min low-intensity (55-70% HR max ) cool-down. Sessions were performed at the same time of day throughout the intervention period with room temperature maintained at C and 50-60% relative humidity. Subjects manipulated cycling load electronically by adjusting the ergometer with ±3 W precision, and were provided with continuous feedback regarding their absolute and average power, cadence (RPM), HR, and elapsed time on a large video screen. RPM was individually selected. During interval sessions, subjects were instructed to cycle at their maximal sustainable intensity during all four interval bouts (isoeffort) (28, 29) such that they: (1) completed the described session structure (all four interval bouts completed with only 2 min rest), and (2) with even or progressive power from 1st to 4th interval bout. Prior to each interval session, we estimated the power each subject would be able to maintain during all interval bouts based on previous interval sessions and subject feedback. power, HR (mean and peak), rating of perceived exertion (RPE) 6-20 (3) and RPM were quantified at the end of each interval lap. Blood lactate concentration [la - ] was measured randomly among a subset of 56 subjects at the end of the 3 rd and 4 th interval bout. Data from all intervention groups pooled together showed that the three different interval prescriptions (4x16 min, 4x8 min and 4x4 min) induced significantly different mean power, [la - ], and HR (mean and max) responses. In addition, both RPE and srpe (9), were significantly

4 different across interval prescriptions despite the same maximal session effort approach (Table 1). However, all intervention groups (INC, DEC and MIX) executed the three different interval prescriptions with similar mean power, [la - ], HR (mean and max), RPE and srpe. In addition, there was no significant difference in total compliance (% interval sessions completed) among intervention groups. TABLE 1. Physiological and perceptual responses during interval sessions executed as 4x16, 4x8 and 4x4 min during a 12-wk intervention period. 4x16 min 4x8 min 4x4 min P-value* Compliance (% HIT sessions) 93.1 (14.2) 96.4 (8.8) 92.5 (13.2) Power (W) 276 (25) 308 (29) 342 (33) <0.001 Power (W. kg -1 ) 3.5 (0.4) 3.9 (0.4) 4.3 (0.4) <0.001 Percent of Peak Power Output (%) 65 (4) 71 (4) 80 (4) <0.001 Percent of 4mM lactate power (%) 97 (8) 106 (8) 118 (9) <0.001 Blood lactate (mmol. L -1 ) # 4.7 (1.6) 9.2 (2.4) 12.7 (2.7) <0.001 Interval lap HR mean (% HR peak ) 86 (3) 88 (2) 89 (2) <0.001 Interval lap HR peak (% HR peak ) 89 (2) 91 (2) 94 (2) <0.001 RPE (6-20) 15.0 (1.1) 16.2 (0.8) 17.1 (0.9) <0.001 srpe 30min post session (1-10) 6.3 (1.0) 6.9 (1.0) 7.7 (1.2) <0.001 All values are calculated as the mean of means (SD) of up to 24 training sessions in 63 subjects. Compliance is calculated as percent of total interval sessions executed in relation to number of described sessions (24 in each subject). All values of power, mean/peak heart rate (HR) and rate of perceived exertion (RPE) represent a mean of all 4 interval laps. Session RPE (srpe) was quantified 30 min post exercise. # Blood lactate was measured randomly among a subset of 56 subjects after interval lap 3 and 4, and a total of 531 samples (~10 per participant) were collected. * One way repeated measure ANOVA. Training monitoring All subjects were provided with the Norwegian Olympic committee s online training diary to record their training. The following variables were registered for each training session: (1) total training form duration (endurance, strength, sprint/jump, other), (2) activity form duration (cycling, running, XC skiing etc.), (3) total duration in each endurance training zone (session goal/time in zone-method (33)), (4), session goal (SG) categorical intensity distribution (33), (5) perceived exertion (1-10) rated 30 min post-exercise (srpe) (8) and (6) self-reported recovery status (1-9) (19). Individualized heart rate (HR) zones were calculated based on HR peak results from pre-testing using a 5-zone aerobic intensity scale utilized by the Norwegian Olympic Federation to prescribe and monitor training of well-trained endurance athletes: zone % HR peak, zone % HR peak, zone % HR peak, zone % HR peak, zone % HR peak (29). There were no significant differences among groups in any training variable measured as mean during 12-wk (Table 2), and no significant differences in training volume during the intervention period compared to the previous training year. Weekly training volume remained stable across mesocycle 1-3 in all groups (average cycle 1: 9.8 ± 3.2 h. wk -1, cycle 2: 10.0 ± 3.2 h. wk -1, cycle 3: 10.7 ± 3.1 h. wk -1 ). A self-reported scale for recovery status suggested that subjects were fully recovered every 4 th wk, as there were no significant differences among the three intervention groups or across 4-wk training cycles in self-reported recovery status (data not shown).

5 TABLE 2: Weekly training characteristics and sickness during a 12-wk training period in 63 subjects, randomized to increasing HIT (INC), decreasing HIT (DEC) and mixed HIT (MIX) training groups. Values are mean (SD). All INC DEC MIX P-value* (N=63) (N=23) (N=20) (N=20) Training volume (h. wk -1 ) 10.1 (2.9) 10.8 (2.6) 9.9 (3.1) 9.6 (2.9) Training forms Endurance (%) 96.9 (3.7) 97.2 (4.2) 96.6 (3.3) 97.0 (3.7) Strength (%) 2.6 (3.5) 2.3 (4.1) 2.7 (3.2) 2.7 (3.1) Speed/jumps (%) 0.1 (0.3) 0.0 (0.1) 0.2 (0.4) 0.0 (0.1) Other (%) 0.4 (0.9) 0.4 (0.9) 0.5 (0.9) 0.3 (0.8) Intensity distribution Zone 1 (%) 71.2 (13.7) 72.8 (12.5) 67.7 (15.0) 72.8 (13.7) Zone 2 (%) 12.3 (9.0) 11.6 (8.3) 15.9 (9.8) 9.4 (8.1) Zone 3 (%) 8.9 (3.8) 9.0 (3.5) 8.4 (3.5) 9.4 (4.6) Zone 4 (%) 5.3 (2.5) 4.7 (1.8) 5.3 (2.5) 5.9 (3.0) Zone 5 (%) 2.3 (1.4) 1.9 (1.0) 2.7 (1.5) 2.5 (1.7) Specific training (%) 81.3 (15.1) 78.0 (17.8) 84.0 (14.0) 82.5 (12.6) Sickness (days) 3.8 (3.6) 3.1 (2.4) 3.1 (3.1) 5.2 (4.7) Intensity distribution and specific training are calculated as percent of endurance training, and distributed according to session goal/time in zone-method (SG/TIZ) (33). Zone 1 = 60-75% of HR peak, zone 2 = 75-85% of HR peak, zone 3 = 85-90% of HR peak, zone 4 = 90-95% of HR peak, zone 5 = % of HR peak. *One-way between-groups ANOVA. Testing procedures Pre-testing was completed two wk before intervention start. Post-testing was initiated 2-4 days after the last supervised interval session for all subjects, and completed within 10 days. Both testing periods were performed over two days separated by a minimum of 48 h recovery. Subjects were instructed to perform only LIT for a minimum of 48 h preceding each test and to consume the same type of meal. They were instructed to not eat during the last hour, or consume caffeine during the last 3 h preceding testing. Test day 1 On day 1, 4-6 submaximal incremental 5-min steps were performed in the laboratory on a bicycle ergometer to identify the workload eliciting 4 mmol. L -1 [la - ] (Power 4mM ) and gross efficiency (GE). The test started with 5 min cycling at 125 W, and O 2, respiratory exchange ratio (RER), and HR were measured during the last 2.5 min, with mean values for this period used for statistical analyses. Blood [la - ] was measured after 4.30 min, and RPE was determined at the end of each 5-min step using Borg s 6-20 RPE scale (3). Power was increased by 50 W (25 W if [la - ] was >3 mmol. L -1 ) after 5 min. Testing was terminated when [la - ] reached 4 mmol. L -1. Power and O 2 corresponding to 4 mmol. L -1 [la - ] were identified after plotting the true power-lactate curve for each subject, by fitting a polynomial regression model (18). GE was calculated using the method of Coyle et al. (5). Briefly, rate of energy expenditure was calculated by using gross O 2 from the first three 5-min submaximal steps (125, 175 and 225 W), and GE was expressed as the ratio of work accomplished per minute to caloric expenditure per minute.

6 After 10 min recovery, an incremental test to exhaustion was performed to determine: (1) O 2peak, (2) peak power output (PPO), (3) HR peak, and (4) peak blood lactate concentration [la - peak]. The test started with 1 min of cycling at 3 W. kg -1 (rounded down to nearest 50 W), and subsequently increased by 25 W every minute until voluntary exhaustion or failure to maintain 70 RPM. Strong verbal encouragement was provided throughout the test. O 2peak was calculated as the average of the two highest 30-sec consecutive O 2 measurements. Plateau of O 2 curve and/or HR 95% of known HR max, RER 1.10 and [la - ] 8.0 mmol. L -1 were used as criteria for the attainment of O 2peak (10). PPO was calculated as the mean power during the last minute of the test. HR peak was recorded during the final 5 sec before exhaustion and [la - peak] was measured 60 sec post-exhaustion. In addition, a theoretical maximal aerobic power (MAP) was calculated by using submaximal O 2 measurements in addition to O 2peak. MAP was defined as the power where the horizontal line representing O 2peak meets the extrapolated linear regression representing the submaximal O 2 /power relationship. To estimate fractional utilization of O 2peak, the previously described O 2 corresponding to 4 mmol. L -1 [la - ], was calculated as percentage of O 2peak ( O Finally, after 15 min recovery, a 30 s all-out Wingate test (38) was conducted. The test started with the subject pedaling at a freely chosen cadence below 120 RPM for 20 s with a ~150 W braking resistance. Then, following a 3 s countdown, a braking resistance equivalent to 0.7 Nm. kg -1 body mass (Lode Excalibur), or a torque factor (Velotron) was applied to the flywheel and remained constant throughout the 30 s test. Cyclists were instructed to pedal as fast as possible from start and were allowed to sit or stand as preferred throughout the test. Strong verbal encouragement was provided throughout. power during 30 s (Power 30s ) was recorded. Test day 2 On test day 2, subjects performed a 40-min all-out trial (Power 40min ). The test started with a 30-min warm-up at a self-selected power output. Thereafter, subjects were instructed to cycle at the highest possible mean power during 40-min. Subjects were blinded to power output and HR, but were allowed to see remaining time and RPM. They were encouraged to remain seated during the trial, but were permitted to stand and stretch their legs occasionally, and were allowed to drink water ad libitum. power, mean HR (HR mean ) and HR peak were registered, as well as RPE and [la - ] at the end of the test. Instruments and materials For each individual, all tests on day 1 were performed on the same Velotron (Racermate, Seattle, WA, USA) or Lode Excalibur Sport (Lode B. V., Groningen, The Netherlands) cycle ergometer under similar environmental conditions (18-22 C/50-60% relative humidity). Pre- and post-tests were performed at the same time of day. Saddle height, handlebar position and distance between tip of the saddle and the bottom bracket were adjusted by each subject as desired. Subjects were instructed to remain seated during all tests (with the exception of the 30 s all-out test) and allowed to choose their preferred cadence. Both test ergometers are computer controlled and provide <2% margin of error in both accuracy and repeatability, according to the manufacturer. Test day 2 and all interval sessions were performed in groups on their own road racing bicycle mounted on Computrainer Lab TM ergometers (Race Mate, Seattle, WA, USA) calibrated according to the manufacturer s specifications and connected to a central PC running dedicated software (PerfPRO Studio, Hartware Technologies). O 2 was measured using Oxycon Pro with mixing chamber and 30 s sampling time (Oxycon, Jaeger GmbH, Hoechberg, Germany). Gas sensors were calibrated via an automated process using certified calibration gases of known concentrations before every test. The flow turbine (Triple V, Erich Jaeger) was calibrated using a 3L calibration syringe (5530 series; Hans Rudolph, Kansas, MO, USA). HR was measured using Polar V800 (Polar Elektro Oy, Kempele, Finland). Blood [la - ] were analyzed using a stationary lactate analyzer (EKF BIOSEN, EKF diagnostic, Cardiff, UK).

7 Statistical analyses Data were analyzed using SPSS 22.0 (SPSS Inc, Chicago, IL, USA) and are presented as mean ± standard deviation (SD) or 95% confidence intervals. Baseline and training characteristics were compared using a one-way between-groups analysis of variance (ANOVA), followed by Bonferroni corrected post hoc tests. A one-way repeated measures ANOVA was used to compare differences among 4x16 min, 4x8 min and 4x4 min interval session prescriptions. A univariate General Linear Model (GLM) (analysis of covariance (ANCOVA)) was used to assess differences in baseline characteristics and changes in test variables among the intervention groups. A GLM repeated measures model (ANOVA) was used to compare pre- and post-test results in each group. GLM analyses were adjusted for the influence of different covariates (test-location and pre Power 4mM (w. kg - 1 )), and conducted to ensure that there were no violations of the assumptions of normality, linearity and sphericity. All data analyzed by GLM are presented as adjusted values. Due to expectations of small changes in these already well-trained cyclists, the data were further analyzed with effect size (ES) calculated according to Cohen s d (0.2=small, 0.5=medium, 0.8=large) (4). Medium or large ES (>0.5) are discussed as tendencies if comparisons are non-significant. The frequency distribution of individual response magnitude across training groups was compared using a Chi square test, and ES was calculated with Cramer s V with three categories (4). For all comparisons, statistical significance was accepted as α RESULTS Baseline characteristics and body mass There were no significant differences among training groups before the intervention period with respect to age, cycling experience, body mass, or any performance or physiological test variables (Table 3). After the intervention, there was a significant body mass reduction in INC (80.3±7.4 vs. 79.0±7.6 kg), DEC (79.7±7.8 vs. 78.5±7.5 kg) and MIX (79.7±8.9 vs. 78.2±8.8 kg) training groups (all P<0.05). Performance responses All training groups improved significantly in all performance measures after the intervention period. improvement pre-post in Power 40min was 8.0 (5.3, 10.6), 7.4 (4.4, 10.4) and 4.9% (1.8, 8.0) in INC, DEC and MIX group, respectively (all P<0.05; Figure 2). The relative improvement did not differ among groups (P=0.307), but there was a medium ES when comparing difference in absolute values (Table 3) in INC and DEC vs. MIX groups. PPO increased significantly by 7.1 (4.7, 9.5), 6.0 (3.4, 8.6) and 6.5% (3.9, 9.2) in INC, DEC and MIX group, respectively (all P<0.05; Figure 2), with no differences among groups (P=0.813). MIX and DEC groups improved significantly in mean Power 30s by 2.4 (0.3, 4.4) and 2.7% (0.7, 4.7), respectively (both P<0.05), while a non-significant 1.2% (-0.7, 3.1) change occurred in the INC group. The changes in Power 30s did not differ among groups (P=0.509). Physiological responses The INC and DEC groups improved mean Power 4mM significantly by 5.8 (2.7, 8.9) and 5.9% (2.6, 9.2), respectively (all P<0.05). MIX group showed a non-significant change of 2.9% (-0.4, 6.3) (Figure 2). The relative changes among groups in Power 4mM did not differ (P=0.360), but there was a medium ES when comparing absolute values (Table 3) in INC vs. MIX group. All groups significantly improved mean O 2peak by 5.8 (3.7, 8.0), 4.5 (2.3, 6.8) and 3.8% (1.5, 6.0) in INC, DEC and MIX groups, respectively (all P<0.05; Figure 2). No significant differences occurred among groups (P=0.392), but there was a medium ES when comparing absolute values (Table 3) in INC vs. MIX group. DEC group significantly improved mean fractional utilization calculated as O by 3.7% (1.2, 6.3) (P<0.05). There was a non-significant 1.3 (-1.1, 3.7) and -0.5% (-3.1, 2.1) change in INC and MIX groups, respectively (Figure 2). Although the relative changes among groups did not differ (P=0.070), there was a medium ES when comparing DEC vs. MIX group. All groups decreased in GE. relative changes were -2.6 (-4.4, -0.9) in INC (P<0.05), -2.0 (-3.8, -0.2) in

8 DEC (P<0.05) and -1.4% (-3.3, 0.4) in MIX group (not significant) (Figure 2), with no significant differences among groups (P=0.642). FIGURE 2. 95% CI for relative change after a 12-wk training period (PRE-POST ) in (A) Power 40min, (B) PPO (C) Power 4mM, (D) O 2peak, (E) % O and (F) GE, in Increasing HIT (INC) (N=23), Decreasing HIT (DEC) (N=20) and Mixed HIT (MIX) (N=20) intervention groups. Power 40min = power during a 40-min all-out trial, PPO = Peak Power Output, Power 4mM = Power corresponding to 4mMol. L -1 lactate, O 2peak = Peak oxygen uptake, % O = Percent peak oxygen uptake corresponding to 4mMol. L -1 lactate, GE = Gross Efficiency.

9 TABLE 3. PRE-values and PRE to POST changes in performance and physiological variables during a 12-wk training period with different periodization models in increasing HIT (INC), decreasing HIT (DEC) and mixed HIT (MIX) training groups. All values are mean. ALL GROUPS (N=63) PRE Body composition Body mass (kg) 79.7 (77.9, 81.5) Performance Power 40min (W) 281 (274, 288) PPO (W) 413 (406, 421) Aerobic Power 4mM (W) 281 (275, 288) O 2peak (ml. min -1 ) 4858 (4742, 4974) O 2peak (ml. kg -1. min -1 ) 61.3 (60.1, 62.4) MAP (W) 371 (362, 381) change -1.3* (-1.7, -0.9) 19* (14, 23) 26* (20, 31) 13* (8, 18) 226* (163, 288) 3.9* (3.1, 4.7) 12* (6, 19) PRE 80.3 (76.9, 83.6) 281 (267, 295) 416 (400, 431) 276 (265, 287) 4941 (4736, 5146) 61.8 (59.5, 64.1) 376 (361, 390) INC (N=23) change -1.3* (-1.9, -0.7) 23* (14, 32) 30* (19, 41) 17* (6, 28) 299* (191, 407) 4.8* (3.5, 6.1) 19* (6, 32) PRE 79.5 (76.6, 82.4) 279 (269, 289) 414 (400, 427) 283 (273, 292) 4793 (4585, 5002) 60.6 (58.7, 62.5) 372 (355, 388) DEC (N=20) change -1.2* (-2.1, -0.4) 19* (10, 27) 22* (14, 31) 14* (7, 22) 197* (103, 290) 3.5* (2.0, 5.0) 3 (-8, 15) PRE 79.4 (75.6, 83.2) 287 (275, 299) 413 (400, 426) 286 (272, 300) 4859 (4631, 5088) 61.6 (59.8, 63.4) 369 (348, 390) MIX (N=20) change -1.6* (-2.4, -0.8) 10* (4, 16) 25* (12, 37) 6 (-5, 16) 140* (4, 276) 3.0* (1.3, 4.7) 12 (0, 25) Among groups - relative change P-value a Effect size b INC/DEC vs. MIX / / / / / / /-0.4 % O (%) 79.2 (77.9, 80.4) 1.1 (-0.1, 2.3) 77.3 (74.7, 80.0) 0.7 (-1.4, 2.8) 79.4 (77.3, 81.5) 2.8* (0.6, 5.1) 80.7 (78.7, 82.7) -0.4 (-2.5, 1.7) /0.7 GE (%) * * * /0.3 (18.8, 19.3) (-0.6, -0.2) (18.4, 19.3) (-0.9, -0.2) (18.9, 19.7) (-0.7, -0.1) (18.7, 19.5) (-0.7, 0.2) Anaerobic Power 30s (W) 826 (809, 842) 16* (7, 25) 849 (825, 873) 10 (-6, 26) 820 (789, 851) 20* (3, 36) 812 (778, 845) 18* (1, 36) /0.1 Power 40min = power during 40-min all-out trial, PPO = Peak Power Output, Power 4mM = Power corresponding to 4mMol. L -1 lactate, O 2peak = Peak oxygen uptake, % O = Percent peak oxygen uptake corresponding to 4mMol. L -1 lactate, GE = Gross Efficiency, Power30s = power during 30 s all out test. *P<0.05 PRE vs POST within group, a General Linear Model univariate, adjusted for test-location and pre power at 4mMol. L -1 lactate (w. kg - 1 ). b Effect size calculations according to Cohen s d (0.2=small, 0.5=medium, 0.8=large) (4)

10 A Chi-square test for independence indicated no significant association among training groups and individual performance (Power 40min ) response (P=0.146, Figure 3). There was, however, a medium ES (4), calculated with Cramer s V with three categories. ~87, 63 and 56% of subjects in INC, DEC and MIX group, respectively, achieved moderate to large gains in performance capacity, while ~13, 37 and 44% showed non-response. FIGURE 3. Individual response distribution to PRE-POST change (%) in performance (mean power during 40-min all-out trial) after 12-wk training in Increasing HIT (INC) (N=23), Decreasing HIT (DEC) (N=19) and Mixed HIT (MIX) (N=18) intervention groups. Percent change was categorized as non-response: <3% change, moderate response: 3-9% change, or large response: >9% change. DISCUSSION The present study demonstrates that, at the group level, the physiological and performance improvements following intensified training were moderate to large in all training groups utilized in this study. This indicates that the basic load features of the training were well tolerated and effective. However, the specific HIT periodized mesocycle order or mixed distribution, focusing on manipulating the intensity prescription for interval sessions, had little or no generalizable effect on the adaptive impact of the same overall endurance training load. Furthermore, the individual variation in training response did not significantly differ among the three training groups, suggesting similar expected distribution of large, moderate or non-responses, respectively, to each prescription. Performance and physiological adaptations Following a 12-wk training period, including 2-3 interval sessions each wk in addition to ad libitum LIT, we found that all groups significantly increased performance variables (Power 40min and PPO) by 5-8%. Coinciding with 40-min all-out trial improvements, Power 4mM also increased by 3-6% in all groups. These performance response magnitudes are consistent with previous studies investigating the effect of HIT over similar time frames (14, 19, 25), or after shorter HIT interventions (2-6% improvement) (12, 32). Furthermore, all groups increased VO 2peak significantly by 4-6%, which is in line with the increase in VO 2max reported in other studies involving well-trained to elite level cyclists during comparable training periods (14, 19, 25). Overall, our results demonstrate that the training load prescribed in the present study was effective in improving performance and physiological capacity in well-trained cyclists. We found negligible changes in the fractional utilization of VO 2peak from pre to post test, both in INC (~1%) and MIX (~0%) group. The overall small changes in this variable are likely due to the fact that short-term HIT stimuli are more effective in inducing central cardiovascular adaptations (12). However, DEC group improved by ~4%. A small decrease in GE occurred in all groups, despite increased VO 2peak. A relative shift in energetic contribution from carbohydrate to fat could account for a small decrease in GE. For example, a shift in RER from 0.87 to 0.82 at the same oxygen consumption and power output would result in a ~1% decline in GE (from for example, 21.6 to 21.4%). However, the decrease in GE observed in the present

11 study was still larger than what could be explained by a shift in RER towards greater fat utilization. The main contributor to decreased GE is therefore probably due to higher oxygen consumption, which has also been reported previously (9). Group comparisons Despite large overall progress in all groups, we found no significant differences among groups in adaptive changes from pre to post intervention with the exception of fractional utilization of VO 2peak where DEC group tended to improve more than the other groups. The latter may be a compensation of the slightly smaller increase in VO 2peak in DEC compared to INC group. Altogether, these results suggest that organizing different interval sessions in a specific periodized increasing or decreasing mesocycle order, or in a mixed intensity distribution results in minor differences in adaptive response when the overall load is the same. However, although there were no significant differences among groups, the greater micro-variation of interval training stimuli (i.e., MIX group) tended to induce less overall adaptive responses compared to the INC and DEC group. We speculate that this tendency could be explained by higher interval session quality in the INC and DEC groups who, unlike MIX group, performed the same eight interval sessions consecutively during each mesocycle. Therefore, subjects in the INC and DEC groups may have been more familiar with their specific sessions, and thus able to more accurately pace their tolerable power/intensity from the beginning of the 1 st to the end of the 4 th interval-bout. We have failed to find any experimental studies for direct comparisons with our results. However, previous experimental studies manipulating HIT organization patterns during timeframes from 2-12 wk, indicate improved block periodization training adaptations compared to mixed programs (19-21) and superior effects following a polarized TID compared to a HIT block periodization training concept (30). However, in these studies, block periodization was organized as short periods with heavy HIT stimulus followed by periods with LIT focus, or without same total training load among groups, and is therefore not directly comparable to the present study. Individual differences in adaptions response Despite excellent overall control of the training program variables, and no differences among groups in overall training load, we quantified large individual differences in adaptive response after 12 wk of training. This finding is consistent with other recent studies (15, 36). Furthermore, a response distribution analysis for Power 40min revealed no significant differences in the variability of response across groups (Figure 3). However, we do note that only 56 and 63% of subjects in the MIX and DEC groups achieved >3% improvement, as compared to 87% of subjects in the INC group. Supplementary analyses of variables influencing the individual effects following different periodization models are needed in future studies. Methodological considerations The main strengths of this study were the structured randomized design, rigorous monitoring of all training variables and the large group of well-trained endurance athletes. We managed to match the groups for total work (isoenergetic) and all subjects, regardless of group, performed a welldocumented training model with 2-3 weekly interval sessions interspersed with ad libitum LIT. Based on previous studies using the same model of interval training prescription (28), we anticipated that the different interval duration prescriptions (4x16, 8 and 4 min) would constrain three reasonably discrete work intensities, which would allow us to compare the effects on endurance adaptions when organizing those interval-training prescriptions in different periodized mesocycle groups. The distinctive physiological responses to the three interval prescriptions were confirmed by the significant differences in power, [la - ], HR, RPE and srpe during interval sessions. This study was conducted as a multicentre trial involving three test-locations, which administrated 29, 20 and 20 subjects each, respectively. We are conscious that, despite our best efforts to standardize them, there could be small methodological differences across centres that may affect the intervention results.

12 CONCLUSIONS The present study suggests that organizing different interval sessions in a specific periodized mesocycle order or in a mixed distribution during a 12-wk training period has little or no effect on training adaptation when the overall training load is the same. Although we found a small tendency indicating that a larger micro-variation in interval training intensity and duration (i.e., MIX group) actually induces less adaptation, we overall argue that rigid periodization structures are not supported by the results of this direct intervention study. RECOMMENDATIONS We argue that organizing HIT training in a specific pattern during the preparation period leading up to the competition period, has little or no effect on athletes competing in middle or long-distance running. We recommend to maintain a high total training volume and an appropriate intensity distribution as long as possible until the final tapering period before the most important competitions. The findings of this study suggest that general training patterns are more important than periodization of HIT sessions. REFERENCES 1. Billat V, Lepretre PM, Heugas AM, Laurence MH, Salim D, Koralsztein JP. Training and bioenergetic characteristics in elite male and female Kenyan runners. Medicine and science in sports and exercise. 2003;35(2): ; discussion Billat VL, Demarle A, Slawinski J, Paiva M, Koralsztein JP. Physical and training characteristics of top-class marathon runners. Medicine and science in sports and exercise. 2001;33(12): Borg G. Perceived exertion as an indicator of somatic stress. Scandinavian journal of rehabilitation medicine. 1970;2(2): Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale (NJ): Lawrence Erlbaum Associates; p. 5. Coyle EF, Sidossis LS, Horowitz JF, Beltz JD. Cycling efficiency is related to the percentage of type I muscle fibers. Medicine and science in sports and exercise. 1992;24(7): De Pauw K, Roelands B, Cheung SS, de Geus B, Rietjens G, Meeusen R. Guidelines to classify subject groups in sport-science research. International journal of sports physiology and performance. 2013;8(2): Esteve-Lanao J, Foster C, Seiler S, Lucia A. Impact of training intensity distribution on performance in endurance athletes. Journal of strength and conditioning research / National Strength & Conditioning Association. 2007;21(3): Foster C, Florhaug JA, Franklin J et al. A new approach to monitoring exercise training. Journal of strength and conditioning research / National Strength & Conditioning Association. 2001;15(1): Hopker J, Coleman D, Jobson SA, Passfield L. Inverse relationship between V O2max and gross efficiency. International journal of sports medicine. 2012;33(10): Howley ET, Bassett DR, Jr., Welch HG. Criteria for maximal oxygen uptake: review and commentary. Medicine and science in sports and exercise. 1995;27(9): Jeukendrup AE, Craig NP, Hawley JA. The bioenergetics of World Class Cycling. Journal of science and medicine in sport / Sports Medicine Australia. 2000;3(4): Laursen PB, Shing CM, Peake JM, Coombes JS, Jenkins DG. Interval training program optimization in highly trained endurance cyclists. Medicine and science in sports and exercise. 2002;34(11): Lucia A, Hoyos J, Pardo J, Chicharro JL. Metabolic and neuromuscular adaptations to endurance training in professional cyclists: a longitudinal study. The Japanese journal of physiology. 2000;50(3): Lucia A, Hoyos J, Perez M, Chicharro JL. Heart rate and performance parameters in elite cyclists: a longitudinal study. Medicine and science in sports and exercise. 2000;32(10):

13 15. Mann TN, Lamberts RP, Lambert MI. High responders and low responders: factors associated with individual variation in response to standardized training. Sports Med. 2014;44(8): Issurin VB. New Horizons for the Methodology and Physiology of Training Periodization. Sports Medicine. 2010;40(3): Neal CM, Hunter AM, Brennan L et al. Six weeks of a polarized training-intensity distribution leads to greater physiological and performance adaptations than a threshold model in trained cyclists. Journal of applied physiology. 2013;114(4): Newell J, Higgins D, Madden N et al. Software for calculating blood lactate endurance markers. Journal of sports sciences. 2007;25(12): Rønnestad BR, Ellefsen S, Nygaard H et al. Effects of 12 weeks of block periodization on performance and performance indices in well-trained cyclists. Scandinavian journal of medicine & science in sports. 2014;24(2): Rønnestad BR, Hansen J, Ellefsen S. Block periodization of high-intensity aerobic intervals provides superior training effects in trained cyclists. Scandinavian journal of medicine & science in sports. 2014;24(1): Rønnestad BR, Hansen J, Thyli V, Bakken TA, Sandbakk O. 5-week block periodization increases aerobic power in elite cross-country skiers. Scandinavian journal of medicine & science in sports. 2015;26(2): Sandbakk O, Hegge AM, Losnegard T, Skattebo O, Tønnessen E, Holmberg HC. The Physiological Capacity of the World's Highest Ranked Female Cross-country Skiers. Medicine and science in sports and exercise Sandbakk O, Holmberg HC, Leirdal S, Ettema G. The physiology of world-class sprint skiers. Scandinavian journal of medicine & science in sports. 2011;21(6):e Sandbakk O, Sandbakk SB, Ettema G, Welde B. Effects of intensity and duration in aerobic high-intensity interval training in highly trained junior cross-country skiers. Journal of strength and conditioning research / National Strength & Conditioning Association. 2013;27(7): Sassi A, Impellizzeri FM, Morelli A, Menaspa P, Rampinini E. Seasonal changes in aerobic fitness indices in elite cyclists. Applied physiology, nutrition, and metabolism = Physiologie appliquee, nutrition et metabolisme. 2008;33(4): Schumacher YO, Mueller P. The 4000-m team pursuit cycling world record: theoretical and practical aspects. Medicine and science in sports and exercise. 2002;34(6): Seiler S, Hetlelid KJ. The impact of rest duration on work intensity and RPE during interval training. Medicine and science in sports and exercise. 2005;37(9): Seiler S, Joranson K, Olesen BV, Hetlelid KJ. Adaptations to aerobic interval training: interactive effects of exercise intensity and total work duration. Scandinavian journal of medicine & science in sports. 2013;23(1): Seiler S, Tønnessen E. Intervals, thresholds, and long slow distance: the role of intensity and duration in endurance training. Sportscience. 2009;13: Stoggl T, Sperlich B. Polarized training has greater impact on key endurance variables than threshold, high intensity, or high volume training. Frontiers in physiology. 2014;5: Stöggl TL & Sperlich B. The training intensity distribution among well-trained and elite endurance athletes. Frontiers in physiology. 2015;6: Swart J, Lamberts RP, Derman W, Lambert MI. Effects of high-intensity training by heart rate or power in well-trained cyclists. Journal of strength and conditioning research / National Strength & Conditioning Association. 2009;23(2): Sylta O, Tonnessen E, Seiler S. From heart-rate data to training quantification: a comparison of 3 methods of training-intensity analysis. International journal of sports physiology and performance. 2014;9(1): Tønnessen E, Svendsen IS, Rønnestad BR, Hisdal J, Haugen TA, Seiler S. The annual training periodization of 8 world champions in orienteering. International journal of sports physiology and performance. 2015;10(1):29-38.

14 35. Tønnessen E, Sylta O, Haugen TA, Hem E, Svendsen IS, Seiler S. The road to gold: training and peaking characteristics in the year prior to a gold medal endurance performance. PloS one. 2014;9(7):e Vesterinen V, Hakkinen K, Laine T, Hynynen E, Mikkola J, Nummela A. Predictors of individual adaptation to high-volume or high-intensity endurance training in recreational endurance runners. Scandinavian journal of medicine & science in sports Aug 6. doi: /sms [Epub ahead of print] 37. Zapico AG, Calderon FJ, Benito PJ et al. Evolution of physiological and haematological parameters with training load in elite male road cyclists: a longitudinal study. The Journal of sports medicine and physical fitness. 2007;47(2): Zupan MF, Arata AW, Dawson LH et al. Wingate anaerobic test peak power and anaerobic capacity classifications for men and women intercollegiate athletes. Journal of Strength and Conditioning Research. 2009;23(9):